Search papers, labs, and topics across Lattice.
This study employs an activation-guided intervention framework on the Qwen3-8B model to explore how specific neuron activations associated with Alzheimer's disease (AD) can influence language generation. By amplifying or attenuating the output of neurons identified through their activation rates in AD-related transcripts, the researchers observed significant changes in cognitive performance across various tasks, including story recall and verbal fluency. The results indicate that manipulating these neural activations can effectively induce AD-related computational language phenotypes, highlighting a novel approach to understanding the relationship between language and cognitive dysfunction in AD.
Amplifying neurons linked to Alzheimer's can drastically impair language performance, mirroring cognitive decline seen in patients.
Changes in spontaneous speech provide an early signal of cognitive dysfunction in Alzheimer's disease (AD) that large language models (LLMs) can detect. However, detection alone cannot establish whether the underlying model representations contribute functionally to behavior. We introduce an activation-guided intervention framework using Qwen3-8B. The framework identifies feed-forward neurons with higher activation rates for AD than control transcripts and modulates their output contributions during generation by scaling the corresponding down-projection weights. This yielded nine edited variants differing in intervention direction, magnitude, and scope. The original and edited models completed the same 12-turn neuropsychological battery, assessed through blinded human ratings and computational linguistic measures. Amplifying AD-associated neurons produced graded impairments in story recall, verbal fluency, working memory, procedural discourse, scene construction, and coreference resolution. Attenuation largely preserved performance and selectively improved several outcomes. Amplification also reduced lexical surprisal, idea density, syntactic complexity, and discourse quantity, broadly paralleling changes reported in human AD speech. These findings show that neurons identified solely from clinical language differences can influence behavior across multiple cognitive domains, providing proof of concept for an AD-related computational phenotype and a controlled framework for experimentally examining links between language and broader cognitive dysfunction.
